What is the Impact of AI on Job Market Diversity?

In the past two decades, Artificial Intelligence (AI) and Machine Learning (ML) have made significant advancements and found applications in various fields, ranging from healthcare and environmental monitoring to fraud detection. However, AI has historically struggled to replicate human creativity. Recent breakthroughs, such as the emergence of large language models like ChatGPT, and generative models like Stable Diffusion and DALL-E, have changed this landscape. It is now anticipated that AI will have a profound impact on almost every aspect of our lives, including areas that were previously considered exclusive to human intelligence.

This article aims to explore two key questions: Firstly, how will these new developments in AI impact the job market? Secondly, what are the implications of widespread AI adoption for underrepresented groups? By examining these questions, we can gain insights into the potential opportunities and challenges that arise from the increasing integration of AI into our society.

Throughout human history, new disruptive technologies have dictated the demand for certain skills in labour markets, making certain jobs obsolete but in return introducing many new jobs.  In 2020, the world economic forum predicted that AI will replace up to 85 million jobs worldwide, while at the same time creating 97 million new jobs. The belief at the time was that the jobs lost would largely be those that involved low-skilled repetitive types of work, while new jobs would require strong analytical and problem-solving skills. There has been a distinct shift in these types of predictions since the arrival of generative AI like ChatGPT.  OpenAI’s own report predicts that even higher paying tech jobs are now more likely to be affected. Goldman Sachs predicted that 300 million jobs worldwide will either disappear or be diminished as a result of generative AI. What is striking about these predictions is that perhaps for the first time in history, it is not blue collar work, but white collar office jobs that will be affected by a new technology.  It is however not all doom and gloom - many analysts including ChatGPT’s own Sam Altman predict that there will also be a number of new jobs that will arise as a result of this technology.  It is quite clear that we need to constantly keep reinventing ourselves to be creative and productive, engaged in new job roles, as rapidity with which new skills need to be acquired is on the rise like never before.

It is unclear how these new technologies will affect underrepresented groups when it comes to the job market. However, it is likely that certain jobs like that of a computer programmer will be democratized. The earliest computer programmers needed to have a strong understanding of how every aspect of the computer worked. This changed with the invention of high-level programming languages like Java. Modern computer programmers don’t need to have a detailed understanding of the hardware of the computer – most of them only need a good understanding of programming concepts and fluency in a high-level programming language.  Now, with the arrival of tools such as GitHub Copilot and ChatGPT, there are indications despite challenges that developers can rely, gradually, on tools that automatically produce code, leaving time for them to focus on higher-level tasks. Such a trend may also be seen as a lowering of the bar for entry to a number of software construction jobs, potentially making them more accessible to underrepresented groups. It boils down to product specialists specifying a set of features which can lead to automatic building of the corresponding software component or even an entire system with the aid of 3D printing technology. Humanity has to educate itself in the right ways of using and unleashing the power of technology. 

The predictions should be taken with a huge grain of salt - if history has taught us anything, predictions of how new technologies will change the world are rarely accurate. Consider for example, the dot-com boom. No one in the 1990s could have foreseen the powerful positive and negative impacts of the internet on our lives. 

In conclusion, we must address the potential negative impact of AI on underrepresented groups and minorities. When AI is utilized to make significant decisions that affect individuals' lives, such as in job recruitment, judicial systems, and loan eligibility, it is imperative to ensure that the system is not unfairly biased against any particular group. Unfortunately, bias is often present in AI systems due to the inherent biases in the data they are trained on, reflecting societal prejudices.

Addressing AI fairness has become a critical and active research area. It is crucial to prevent the wide adoption of AI systems from perpetuating or amplifying existing societal biases. Additionally, AI systems exhibit emergent properties, which means that they may display unintended behaviors or biases that were not explicitly programmed. Therefore, it is essential for humans to understand and guide the use of this powerful technology in a safe and reliable manner.

The article has been written by Dr.PVR Murthy, Prof and HOD, Department of Artificial Intelligence and Data Science, Nitte Meenakshi Institute of Technology, Bengaluru